An AI Approach to Teaching Linguistics

Nicholas Sobin


Levin (1983) deals with the problem of teaching linguistics to non-majors. Her major suggestion is that this can be done more successfully by de-emphasizing formalism and by focusing more on areas of greater popular interest, such as sociolinguistics and language variation. Levin's suggestions comprise one very reasonable approach to the problem. What I wish to explore here is an alternative approach, one which uses computation and which actually places greater emphasis on formalism, but in a way which is of interest and benefit to beginning students whether or not they are linguistics majors. Using the artificial intelligence programming language Prolog, one can model competing hypotheses about particular linguistic phenomena and demonstrate the consequences (and hence the desirability) of each hypothesis, AI programming in Prolog gives students a grammar machine to play with. Working with these programs aids in teaching some of the particulars of grammatical analysis, helps illustrate the scientific method of inquiry (formulating alternative hypotheses and assessing their predictions/ consequences), and initiates students into computing and progranuning skills whose benefits far out-extend linguistic applications. 


artificial intelligence; pedagogy

Full Text:



  • There are currently no refbacks.